Led 0→1 Design of a Centralized Identity & Access Platform for NASA Geoscientists

my role

UX Lead — Interview, Synthesis, Ideation, Iteration, Presentation

Team

2 Designers

2 PM (NASA, GT)

5 NASA teams

Time

6 weeks, 2024

Tool

Figma

Status

Beta

Impact

Data was collected from user testing and early implementation

~50%

Reduced setup time

100%

Positive feedback

2000+

Scientists benefited

Background

VEDA is NASA’s open-source platform for exploring and analyzing Earth science data. It's being used by NASA geoscientists and researchers worldwide.

8

8

Applications

12+

12+

Collaborators and Universities

Collaborators and Universities

56+

56+

Publications

challenges - Users

The onboarding and access setup process is complex and inconsistent across VEDA applications.

"Each team has its own setup method—it’s hard to know where to start, especially for new admins."

“We’re adding users by hand through emails. It works, but it doesn’t scale.”

“We’re adding users by hand through emails. It works, but it doesn’t scale.”

challenges - BUSINESS

Leadership recognized that without a unified authentication system, NASA’s vision for scalable, cloud-based science would be difficult to achieve.

solution

MVP of a centralized auth platform that streamlined setup, unified workflows across teams, and enabled scalable user and app management.

Process Highlights

Pain Point 01

Simplifying Application Configuration for Multiple Teams

Now, Each team sets up app access manually using scripts or notebooks—leading to inconsistent workflows and setup errors.

iteration 01

Pressure test with five teams

Flow is correct

Need more configuration sessions

iteration 02

Define technical structure

Break the setup into clear sessions

Align terminology across teams to reduce confusion

Final design

A template-based setup to streamline application creation and role definition

Pain Point 02

Designing a Scalable Group Management System

Users often work across multiple apps but need different access in each. No shared “group” model existed to manage this complexity.

Options

I explored three permission models with teams, PMs and devs

Apps inherit global groups

Apps inherit global groups


Easy for admins to manage

Limited flexibility for app teams

Each app manages its own groups

Gives each app team full control

No shared group for collaboration

Hybrid model

Hybrid model


Balances control with flexibility

Can grow unnecessarily cluttered

Final design

We landed on global groups with per-app roles for simplicity. Future plans include auto-assigning users to groups for easy scalability.

Pain Point 03

Enabling Quick, Temporary Access for External Users

During conferences or workshops, it was difficult to share access—admins had to grant/remove access manually for each guest.

User needs

Invite external collaborators quickly and revoke access easily

Business needs

Expand adoption and reduce friction in outreach use cases

Final design

Shareable group links with permission controls for easy access sharing.

My learnings

#1 Navigating a Complex Problem Space

Understanding a complex technical system requires more than desk research, it means learning the tech and getting close to the users. I used early wireframes to pressure-test ideas and uncover key workflow insights, helping me learn quickly while shaping the design iteratively.

#2 Designing for Now and What Comes Next

Even while working on an MVP, I learned to think long term—balancing immediate needs with scalable foundations. Collaborating with PM to map out future scenarios helped us prioritize smartly and design a system that’s ready to grow.

#3 Ask the Obvious Questions

At first, I hesitated to ask “basic” questions—but one of those ended up sparking the most important conversation in the project. My question about platform choices and whether we needed to build everything ourselves led to a productive debate across the team, and ultimately, a more focused and realistic solution.

Back